Forums / Proposed MRR Fields / Project ID

Project ID

1 posts, 0 answered
  • Dave Marvin
    Dave.Marvin avatar
    13 posts
    Registered:
    02 Nov 2012
    28 Jul 2014
    The proposed "Project ID" attribute broadens the scope of the similar "BPA project number" previously proposed by Nicole; it would provide a link to all projects, including those that are not funded by BPA or otherwise do not have a BPA project code, and include projects conducted prior to BPA's tag distribution activities in the early 2000's.  The function of the Project ID attribute is to link mark/recapture data with a persisted set of descriptors that provide context for the purpose and utility of those data, such as the project's primary (and secondary) objectives, its protocols, a current point of contact, its BPA project number (or other project ID provided by the COE, PNNL, BOR, etc.) and, especially for projects that have run to completion, links to annual/final reports and other project documentation containing study results and analysis.  Ideally, the documentation link would be a permanent digital object identifier (doi) to an electronic document stored in a federal/state/provincial/tribal agency's or academic institution's on-line library, or in a regional repository such as the StreamNet library.  Since some agencies have used the same BPA project number for 25 years, and that project number encompasses many studies, the Project ID value should be separate from, and more specific than, the BPA project number.  The fine folks at the Oak Ridge National Laboratory have an entire Web space dedicated to Data Management for Data Providers, including a section on how to provide documentation and metadata.

    Linking the mark/recapture data with the contextual study descriptors will provide value to researchers and analysts, not involved in the original study, who are interested in reviewing or possibly re-purposing those data for a secondary analysis.  Although cautioned to confer with all data owners before re-using PIT tag mark/recovery data, many analysts find this contact laborious (or difficult, in the case of very old data series) and ignore the warning; this frequently leads to the use of data inappropriate for the analysis, and incorrect results.  While providing a reference to these contextual descriptors won't compel researchers to discuss their analytical assumptions with the original data owners, it should at least help those researchers identify and avoid data that do not meet their own research objectives.

    Link to this post